Google’s Wave announcement yesterday kicked off an orgy of geek ecstasy yesterday. Why not? A novel new interface combining email, instant-messaging, social networking and sharing/collaboration, all backed by Google’s rock-solid platform, and open-sourced to boot. Who couldn’t get excited?
When I first looked at the screenshots and demo of Wave, I got excited too: It’s a software project with big ambitions in several directions at once, and I have a soft spot in my heart for that. But the longer I looked, the more I began thinking, whoa — that is one complex and potentially confusing interface. Geeks will love it, but is this really the right direction for channeling our interactions into software?
One of the most interesting pieces I read this week was this report on a scholarly study of information design comparing the effectiveness of one-column vs. three-column layouts. The focus was more on social-networking sites (Facebook vs. LinkedIn) than on news and reading, but I think the conclusions still hold: People like single-column lists — the interface that Dave Winer calls “the River of News” and that most of us have become familiar with via the rise of the blog.
In Say Everything I trace the rise of this format in the early years of the Web, when designers still thought people wouldn’t know how to, or wouldn’t want to, scroll down a page longer than their screen. It turns out to be a natural and logical way to organize information in a browser. It is not readily embraced by designers who must balance the needs and demands of different groups in an organization fighting for home-page space; and it is the bane of businesspeople who need to sell ads that, by their nature, aim to seduce readers’ attention down paths they didn’t choose. Nonetheless, this study validates what we know from years of experience: it’s far easier to consume a stream of information and make choices about what to read when there’s a single stream than when you’re having to navigate multiple streams.
Wondering why Twitter moved so quickly from the geek precincts into the mainstream? For most users, tweets flow out in a single stream.
I think about all this when I look at the lively but fundamentally inefficient interfaces some news sites are playing with. Look at the Daily Beast’s unbearably cacophonous home page, with a slideshow centerpiece sitting atop five different columns of headlines. There is no way to even begin to make choices in any systematic way or to scan the entirety of the site’s offering. When everything is distracting, nothing is arresting. You must either attend to the first tabloid-red editorial shout that catches your eye — or, as I do, run away.
I feel almost as put off by the convention — popularized by Huffington Post and now increasingly common — of featuring one huge hed and photo and then a jumble of run-on linked headlines underneath. These headlines always seem like orphan captions to me. The assumption behind this design is that you must use the first screen of content to capture the reader’s attention. That’s only the case if you are waving so many things in front of the readers’ eyes in that one screen that you exhaust them.
Google Wave has an open API that will presumably allow developers to remix it for different kinds of users. So just as Twitter’s open API has allowed independent application providers to reconfigure the simple Twitter interface into something far more complex and geeky for those who like that, perhaps Wave will end up allowing users who like “rivers” to take its information in that fashion. But the default Wave looks like a pretty forbidding thicket to navigate.
ELSEWHERE: Harry McCracken wonders whether Wave is “bloatware.”
Post Revisions:
- November 13, 2015 @ 06:26:57 [Current Revision] by Scott Rosenberg
- May 29, 2009 @ 09:10:00 by Scott Rosenberg
Word. Designing a home page is definitely tricky; my organization is setting up a re-do to take place over the next year and I’ll be interested to see how many groups actually make it on to the home page. Internal pages do scroll in a single stream, though!
(Actually, Wave looks like a cluttered up “Outlook Today” screen to me. Meanwhile, Gmail can’t let me see my Tasks in a pane? Argh.)
After having watched the demo, I think the article’s focus is a bit off. Wave isn’t very much like a blog, it’s more of a cross between Outlook and a wiki. Actually, it reminds me of the targets that the Chandler team set for themselves early on, but Wave seems much more likely to me to be an Outlook-killer than Chandler.
As one of the main techy people in my small company, I can tell you that this has tremendous appeal. Just being able to install software on a web server that could then be available to everyone without installing new software on their individual machines would make life much easier. Also, new features and workflows would be much easier to implement and refine over time with a product like Wave because we could roll it out bit by bit. They didn’t get into it, but it’s not hard to imagine a calendaring and task management system that could be integrated with Wave to replicate all the functionality of Exchange server, but with much less cost. And the fact that everything runs natively from a modern browser means that people working offsite could be easily integrated without resorting to things like Outlook Web Access, which is cool, but lacks some of the important features of the rich client.
Especially cool was the idea of having a source control type system for collaboratively edited documents. That’s a major issue for my firm right now and something to manage the process of integrating content produced by a variety of people would be a huge plus for us. All in all, a very exciting idea.
I like your analysis. Wave organizes conversations in hierarchical lists. Programmers love those but ordinary people don’t think that way.
I’m reminded of when Salon.com revamped their web site a few years back and made the big mistake of changing the focus from the single stream of articles. They quickly reverted. I don’t remember if that was during your time their Scott.
Wave has tremendous potential to document workflows and easily allow their creation to be played back as necessary. I see tremendous utility for certain use cases, such as the process leading up to scheduling a group meeting, or being able to trace back accountability for same after the fact. I’ll admit your post does have me wondering whether Wave is overkill or not for this particular use case.
Carlos: the Chandler comparison is interesting. But Chandler was and is really focused on calendaring and task management (with an early focus on email that got dropped) –the whole Outlook-killer thing was always overblown. Interestingly, one of the biggest early goals of Chandler was to do all the stuff Kapor wanted to do without a server, peer-to-peer style. Wave, of course, is precisely the opposite — it can do so many neat tricks because it’s totally hosted, and there is only one master copy of each Wave sitting on the server (Google’s or, in the open-source version, anyone’s, I guess).
Kevin: That would have been the false-start redesign in May 2000 where we tried to force people to click through to “section fronts”/topic pages before getting to story pages. Definitely during my time. I could tell stories! But another time.
In the ancient single-conversation versus branching-conversation divide (which, I agree, usually has non-geeks in the single-thread and geeks in the branching-thread camps) Wave is definitely coming down on the branching side.
I agree, Scott, that Wave has huge potential in the areas you mention. In that sense it’s less an “Outlook killer” than, say, a Basecamp killer.
Email’s great virtue (and also Achilles heel, in re: spam) is that it is fully open and public and it allows you to send messages out of the blue to people you don’t know as long as you have their address. What I’m wondering now — and maybe I missed it in that very long demo — is how (or whether) Wave allows for that kind of communication. If it doesn’t, then its utility is much narrower (though its vulnerability to spam is also much lower).
Scott — thanks for the pointer to the Jack Loechner article. That’s interesting research — particularly as it impacts design for the netbook-size screen.
As someone who works exclusively on a laptop, I’ve noticed how many websites assume their readers are using a much larger screen (could it be because nearly all web designers work on a large screen canvass?). With more and more people using netbooks, though, I’m wondering if more designers will be pressured to finally make their sites work in a smaller frame, which in turn means simplifying the layout more than making everything smaller (you’d hope). And here’s evidence that supports that simpler layout.
What’s the biggest impediment to things going that way? I think you nailed it in suggesting the “demands of different groups in an organization fighting for home-page space.”
Scott, the Google folks didn’t address it specifically, but with a bit of additional software, I think the Wave concept would be highly useful in the calendaring and task management realm. If you had a Wave that was tracking a task or a project, and you could see the status of the various tasks update live as tasks are completed, negotiated, reshuffled, reprioritized, and so on, much tighter coordination becomes possible.
As for your point about email, the demo did address that issue in that yes, Wave is being published as protocol in the same way that email, http, ftp, etc. were all published and made available to all. They called it the three Ps: product, platform, and protocol, with the last being the relevant P.
So Wave is a product that Google will be releasing later this year, but it’s also a platform, which means that you can install a server to handle Waves just like you can install a server to handle email, or web pages, or whatever. Finally, since it’s also a protocol, other people can develop their own implementations of it, and integrate it into their own products. Just like you can have an email client that can support various email protocols, newsgroups, etc., you could create a client that could support the Wave protocol as well.
In the demo, they had a section where they talked about interacting with a couple outside companies that had implemented their own Wave servers and UIs. It operated pretty much like email, in that all they needed was the person’s address and the servers handled the rest.
I just finished reading Dreaming in Code and Google Wave reminds me of the initial Chandler goals. Google Wave is centralized but based on the presentation it is possible to connect servers even different implementations, it is an open protocol. I can imagine it would be possible to do something more p2p on top of this really powerful platform. A Calendar is probably one of the first things people will try to implement. And I believe they are using the right technologies to build it.